Head to head comparison of two currently used nomograms predicting the risk of side specific extra capsular extension to indicate nerve sparing during radical prostatectomy for treatment of prostate cancer [Validation externe des nomogrammes de prédiction de l'extension extra-prostatique spécifiquement au c?té pour des patients atteints de cancer de prostate traités par prostatectomie totale]

摘要

Objective: To perform a head to head comparison of these two nomograms by an external validation combined with an identification of probability cut-offs when to indicate NS. Methods: The full models of the nomograms of Ohori et al. and Steuber et al. were used to calculate the risk of ECE based on PSA and side specific information on clinical stage, biopsy Gleason score, % positive cores, and % cancer in cores. A dataset of 968prostate half lobes was used retrospectively for analysis. All patients underwent laparoscopic robot-assisted or open radical prostatectomy for localized prostate cancer. Results: The predictive accuracy of the Ohori nomogram was at 0.80and for the Steuber Nomogram at 0.78 (comparison P> 0.05). In the calibration plot, the Ohori nomogram showed less departures from ideal predictions than the Steuber nomogram. The best probability cut-off to allow NS for the Ohori nomogram seemed to be ≤ 10%, permitting NS in 59.7% of all cases and being associated with a false negative rate of 10%. The best cut-off for the Steuber nomogram seemed to be ≤ 8%, permitting NS in 44% and associated with a false negative rate 12.5%. Conclusions: The Ohori et al. and the Steuber et al. nomograms allow highly accurate and comparable predictions of the risk of side specific ECE. Level of evidence: 4.